Updated switch/glue patch?

[Adrian Chadd]

On 18 December 2011 02:17, Stefan Bethke <stb at lassitu.de> wrote:

>> Erm, surely that's a bit ridiculous.. surely the locking doesn't need
>> to be that fine grained _and_ multi-levelled. There has to be a better
>> way to do this. :)
>
> Exactly.
>
> I've reimplemented iicbb.c to be slightly more protocol compliant, and to be able to tune transfer speeds to the actual hardware capabilities.  I've timed a single GETSCL (with WITNESS) to 8.7 microseconds, clearly that won't do.

Ok. I'll check it out soon.

> I think I'll look at gpio next, as you have, and see if the overhead can be reduced.

In the meantime, would you mind grabbing the latest code from my git
tree (in the work/ath branch) and basing your work off of that? Even
with delay=10 (the current hardcoded default), it takes up
significantly less CPU than before.

If we can debug and test this locking hacking and my iicbb changes
(which are less intrusive than yours, which I'll review soon and we
can merge in later), I'll push this into -HEAD so we can continue
shaving off cycles and fixing things up.

Thanks again Stefan!


Adrian